WAVEGUIDE AS A COMMUNICATION MKOIUM 



1225 



appear at r('«j;ulai' iiit('r\-als and with sinoollilx- (Iccaxiiit; atnplil udc 'i'liis 

 l)('lia\i()r iiulicatcs thai the major ])<)rli(>ii of Ihc ciici'^y in I lie line was 

 t laA'clliiiji; ill a siiijilc mode, and we deduced that this mode was 'ri']|ii as 

 I'oHows: We ohsei'Ncd that the Nclocily of propatial ion was near thai for 

 the 'ri*]|ii mode hy measiirinji; tlie al)soiule lime Itelwcen pulses, a\'er- 

 aged over many round trijjs. Tliis (^xehided all hut about (» modes wiiose 

 {'ut-off t'reciuencies are near tluit of TEm . Measurement of t I'ansmission 

 loss was made liy observin<2; the rate of (l(>cay of the received pulses 

 a\-eraged over 10 or more round trips. 'Die measured loss was found to 

 he approximately 3 db per mile eompared to a theoretical \'alue of 1.9 

 db ])er mile for TEoi pr()))agation. Il follows that ))i-opai!;atioii must have 



Fig. 13 — Record of pulses after 40 miles of repeated traversal over the 50()-foot 



been taking place in the TEoi mode, for all other modes near TEoi in 

 velocity have theoretical losses well in excess of the observed value. 



To summarize the effects shown in Fig. 12, a great many modes in- 

 cluding TEoi w-ere launched by exciting the waveguide through a small 

 aperture in the end plate. All these modes propagated back and forth in 

 the line for a while, but due to the fact that TEoi has appreciably less 

 loss than the other modes, the energ}^ remaining in the line after a suit- 

 able time delay Avas substantially all in the TEoi mode. This permitted 

 measuring TEoi loss over a distance of many miles by allowing the energy 

 to traverse the 500-foot line many times. 



Fig. 13 records three successive trips of a pulse which had tra\-elled up 

 and down the 500-foot waveguide for a total distance of 40 miles. The 

 pulse shape was still essentially the same as that of the transmitted 

 pulse, although background noise had become clearlj' visible. We cer- 



